Update block docs for: llm.md

docs/content/platform/blocks/update/llm.md

@@ -0,0 +1,136 @@
+
+
+# LLM Blocks Documentation
+
+## AI Structured Response Generator
+
+### What it is
+A block that generates structured responses using Large Language Models (LLMs).
+
+### What it does
+Generates formatted object responses based on given prompts, ensuring the output follows a specific structure or format.
+
+### How it works
+Takes a prompt and expected format, sends it to an LLM, and ensures the response matches the required structure. If the response doesn't match the format, it retries automatically.
+
+### Inputs
+- Prompt: The text prompt to send to the language model
+- Expected Format: Dictionary defining the structure the response should follow
+- Model: Choice of LLM to use (e.g., GPT-4, Claude, etc.)
+- Credentials: API key for the chosen LLM provider
+- System Prompt: Additional context for the model
+- Conversation History: Previous messages for context
+- Retry Count: Number of attempts to get a valid response
+- Prompt Values: Variables to fill in prompt templates
+- Max Tokens: Maximum length of the generated response
+
+### Outputs
+- Response: The structured object generated by the model
+- Error: Any error message if the process fails
+
+### Possible use case
+Extracting specific information from customer reviews into a structured format, such as converting free-text feedback into categorized ratings and comments.
+
+## AI Text Generator
+
+### What it is
+A block that generates free-form text responses using LLMs.
+
+### What it does
+Produces natural language responses based on given prompts without enforcing specific formats.
+
+### How it works
+Sends prompts to an LLM and returns the raw text response, allowing for more creative and flexible outputs.
+
+### Inputs
+- Prompt: The text prompt for the model
+- Model: Choice of LLM to use
+- Credentials: API key for the chosen provider
+- System Prompt: Additional context for the model
+- Retry Count: Number of retry attempts
+- Prompt Values: Variables for template filling
+
+### Outputs
+- Response: The generated text
+- Error: Any error message if the process fails
+
+### Possible use case
+Creating blog post drafts, generating creative stories, or writing marketing copy based on given topics.
+
+## AI Text Summarizer
+
+### What it is
+A block that creates concise summaries of longer texts using LLMs.
+
+### What it does
+Breaks down long texts into manageable chunks, summarizes each chunk, and combines them into a final summary.
+
+### How it works
+Processes text in chunks to handle long documents, maintains context through overlap, and recursively summarizes if needed.
+
+### Inputs
+- Text: The long text to summarize
+- Model: Choice of LLM to use
+- Focus: Specific topic to focus on in the summary
+- Style: Summary format (concise, detailed, bullet points, numbered list)
+- Max Tokens: Maximum length for processing
+- Chunk Overlap: How much context to maintain between chunks
+
+### Outputs
+- Summary: The final summarized text
+- Error: Any error message if the process fails
+
+### Possible use case
+Summarizing long research papers, creating executive summaries of reports, or condensing meeting transcripts.
+
+## AI Conversation Block
+
+### What it is
+A block that manages multi-turn conversations with LLMs.
+
+### What it does
+Handles back-and-forth dialogue between users and AI models, maintaining conversation context.
+
+### How it works
+Processes a list of messages representing a conversation and generates appropriate responses while maintaining context.
+
+### Inputs
+- Messages: List of previous conversation messages
+- Model: Choice of LLM to use
+- Credentials: API key for the chosen provider
+- Max Tokens: Maximum response length
+
+### Outputs
+- Response: The model's reply to the conversation
+- Error: Any error message if the process fails
+
+### Possible use case
+Creating interactive chatbots, virtual assistants, or customer service automation.
+
+## AI List Generator
+
+### What it is
+A block that generates lists of items based on given criteria or source data.
+
+### What it does
+Creates structured lists of items either from provided source data or based on specific focus criteria.
+
+### How it works
+Analyzes source data or focus requirements and generates a properly formatted list of relevant items.
+
+### Inputs
+- Focus: The specific topic or criteria for the list
+- Source Data: Optional text to extract list items from
+- Model: Choice of LLM to use
+- Credentials: API key for the chosen provider
+- Max Retries: Number of attempts to generate a valid list
+- Max Tokens: Maximum response length
+
+### Outputs
+- Generated List: The complete list of items
+- List Item: Individual items from the list
+- Error: Any error message if the process fails
+
+### Possible use case
+Extracting key points from articles, generating todo lists, or creating categorical listings from unstructured text.
+